Geodynamic Signals Detected by Geodetic Methods in Iceland - IfE

Univ.-Prof. Dr.-Ing. Prof. h.c. Günter Seeber anlässlich seines 65. Geburtstagesund der Verabschiedung in den Ruhestand. Wissenschaftliche Arbeiten der FachrichtungGeodäsie und Geoinformatik der Universität Hannover Nr. 258: 39-57, 2006Geodynamic Signals Detected by Geodetic Methods in IcelandPáll Einarsson 1 , Freysteinn Sigmundsson 2 , Erik Sturkell 2 , Þóra Árnadóttir 2 , Rikke Pedersen 2 ,Carolina Pagli 2 , Halldór Geirsson 31 Institute of Earth Sciences, University of Iceland2 Nordic Volcanological Center, Institute of Earth Sciences, University of Iceland3 Icelandic Meteorological Office, ReykjavíkAbstractThe geodynamics laboratory provided by Iceland’s position on an active mid-ocean ridge hasbeen recognized for several decades. Geodetic experiments have been designed and carriedout in Iceland since 1938 to verify various global geodynamic theories, such as Wegener’stheory of continental drift, the sea floor spreading hypothesis, plate tectonics, mantle plumesetc. State-of-the-art techniques have been used to obtain data on crustal displacements withever increasing accuracy to constrain the theories. Triangulation and optical levelling wereused in the beginning, later EDM-trilateration. Network GPS surveying began in 1986 andhas been used extensively since then to study crustal movements. With the addition of InSARand continuous GPS in the last decade we have made a significant stride towards the goal ofgiving a continuous representation of the displacement field in time and space. The largestand most persistent signal is that of the plate movements. Geodetic points in East and WestIceland move with the Eurasia and North America Plates, respectively, and the vectors areconsistent with global models of plate movements. The plate boundary zones are a few tens ofkilometers wide, within which strain accumulates. This strain is released in rifting events orearthquakes that have a characteristic displacement field associated with them. In the Kraflarifting episode in 1975-1984 a 100 km long section of the plate boundary in North Icelandwas affected and divergent movement as large as 8-9 m was measured. The June 2000earthquakes in the South Iceland Seismic Zone were the most significant seismic events in thelast decades. Two magnitude 6.5 earthquakes and several magnitude 5 events were associatedwith strike-slip faulting on several parallel faults along the transform-type plate boundary.Slow post-rifting and post-seismic displacements were detected in the months and yearsfollowing these events, caused by coupling of the elastic part of the crust with the viscoelasticsubstratum. Viscosities in the range 0.3-30 x 10 18 Pa s have been estimated from thetime-decay of these fields. Similar values are obtained from crustal uplift measured aroundthe Vatnajökull glacier due to the reduced load of the glacier in the last century. Magmamovements in the roots of volcanoes are reflected by deformation fields measureable aroundthem. The volcanoes inflate or deflate in response to pressure increase or decrease in magmachambers, and intrusive bodies are revealed by bulging of the crust above them. The mostactive volcanoes in Iceland, Katla, Hekla, and Grímsvötn, appear to be inflating at the presenttime, whereas Krafla and Askja are slowly deflating. An intrusion episode was documentednear the Hengill volcano in 1994-1998 and two intrusion events occurred in theEyjafjallajökull volcano in 1994 and 1999, all of which were accompanied by characteristicdeformation fields.1 IntroductionIceland provides rare opportunities to study a multitude of geodynamic phenomena. It is aplatform of dimensions 300 km x 500 km situated astride a divergent plate boundary and on39

Páll Einarsson et al.top of a hotspot presumed to be fed by a deep mantle plume (Einarsson 1991a, 2001). Inaddition glaciers provide a changing load on the Earth´s crust providing a test of its response(Sigmundsson and Einarsson 1992). Evidence of recent crustal movements are common andwidespread, such as fractures, faults, tilted strata, raised terraces and changing coastlines.Volcanic structures and volcanic events are plentiful. The potential to verify importantgeodynamic hypotheses by direct measurements in Iceland was acknowledged early on, e.g.when a German geodetic-geological expedition came to Iceland in 1938 to verify the Wegenerhypothesis of continental drift (Niemczyk 1943). This effort was interrupted by World War IIbut was later continued when new theories of sea floor spreading and plate tectonics becameestablished in the late sixties and early seventies (e.g. Gerke 1974). German geodesists haveever since played a key role in the study of the kinematics of the crust in Iceland.In this paper we give a short overview of geodynamic processes that are expected to give adetectable signal in geodetic data and give examples that have already been established in therecent literature. For further and deeper treatise on this topic the reader is referred to the bookby Sigmundsson (2006).2 Short History of MethodsThe initial German geodetic effort in 1938 (Niemczyk 1943) was based on conventionalgeodetic techniques, triangulation, optical levelling and gravimetry. Part of the network wasdestroyed during WW II because of suspicion of espionage. The advent of electronic distancemeasurement techniques (EDM) in the sixties allowed more precise measurements and newprojects were initiated by German, USA, and English universities (e.g. Gerke 1974, Torge andDrewes 1977, Decker et al. 1971, Brander et al. 1976). These techniques were applied tostudy the extensive crustal movements that occurred during the Krafla rifting episode in theNorthern Volcanic Zone (Figure 1) that started in 1975 (Tryggvason 1984). Optical leveling,gravity and tilt measurements were also used extensively to monitor plate boundarydeformation and volcano inflation and deflation (Sigurdsson 1980, Tryggvason 1980, Johnsenet al. 1980, Torge 1989). The GPS technology opened up new possibilities and the first GPSprojectwas conducted in 1986. It was a large cooperative project with many participants andthe main target was the South Iceland Seismic Zone (Foulger et al. 1987, 1993). The secondcooperative project was conducted in 1987, mainly around the Northern Volcanic Zone, thistime with a major participation of the University of Hannover. This network was remeasuredin 1990 and 1995 (e.g. Seeber 1989) and provided data for two Doctoral Theses by Jahn(1992) and Völksen (2000). Geodetic GPS-instruments were acquired by Iceland in 1992 andsince then numerous networks have been installed in the active zones, with or withoutparticipation of foreign groups.A large progressive step was taken in 1995-1997 when continuous GPS-stations were set upin Reykjavík and Höfn, straddling the plate boundary. The Continuous GPS-network (CGPS)has been expanding since then (Geirsson et al. 2005 submitted), giving valuable informationon the temporal evolution of the crustal deformation fields. Continuous GPS-stations areoperated at 18 points at the present time providing location every day. The largest build-upphases of the CGPS-network occurred in 1999 in SW-Iceland following the inflation periodof the Hrómundartindur volcanic system near the Hengill triple junction (Sigmundsson et al.1997), in 2000 after the large earthquakes in the South Iceland Seismic Zone, and in 2001around the Katla volcano in S-Iceland following a re-awakening of the volcano in 1999(Sturkell et al. 2003b).40

Geodynamic Signals Detected by Geodetic Methods in IslandInterferometric analysis of synthetic aperture radar images (InSAR) acquired by satellites hasfurther expanded the possibilities to detect geodynamic signals in the surface deformationfield (e.g., Massonnet and Sigmundsson 2000). By comparing interferometrically twosynthetic aperture radar images of the same area, taken at some time interval by the SARsatellites, one can determine surface displacement in the direction of the line-of-sight to thesatellite. If the signals of the two images are sufficiently coherent one obtains a twodimensionalmap of the field. The InSAR technique has been successfully applied in Icelandsince 1997 to a range of geodynamic problems (e.g. Vadon and Sigmundsson 1997,Sigmundsson et al. 1997, Pedersen et al. 2003, Pagli et al. 2005).Figure 1: Index map of Iceland showing place names, lithospere plates, and plate boundary structures (afterEinarsson and Sæmundsson 1987). Divergent plate boundaries are shown with fat, broken lines, seismic zonesand transforms with fat lines. Fissure swarms are shown as gray stripes, central volcanoes with closed, thin lines.The divergent plate boundary zones are the Reykjanes Peninsula, the Western Volcanic Zone (WVZ), theEastern Volcanic Zone (EVZ) and the Northern Volcanic Zone (NVZ). The transform zones are the SouthIceland Seismic Zone (SISZ) and the Tjörnes Fracture Zone (TFZ). The central volcanoes are Hengill andHrómundartindur (H), Hekla, Eyjafjallajökull (E), Katla, Grímsvötn (G), Askja, and Krafla. Vatnajökull andLangjökull are ice caps.The long-term aim of crustal deformation measurements is to obtain a complete, fourdimensionalpicture of the deformation field, i. e. both in space and time. As a step towardthis goal the combined use of network GPS, continuous GPS and InSAR has turned out to bequite successful. In areas where InSAR works, it gives an areal coverage of the deformation41

Páll Einarsson et al.field in one dimension, i.e. in the direction of line-of-sight to the SAR satellite. A network ofGPS-points in the same area allows the determination of the three-dimensional displacementvector, but only at discrete points. The gaps between points can be bridged using InSAR.Continuously recording GPS-stations give the time-history of displacements at a few discretepoints, which is very important when interpreting the processes responsible for the timechanges in the displacement field. Facilities for the use of these methods have beensystematically built up in Iceland in the last decades. The geodetic reference network, ISNET,of about 120 GPS-points covers the country evenly and is resurveyed about every decade(Geirsson et al. 2005). Denser networks are available in the plate boundary zones and aroundmost of the active volcanoes, in most cases with mesh-size of about 5-10 km. Conditions forInSAR analysis of crustal movements are rather favourable in Iceland as lava surfaces androck outcrops in combination with low vegetation provide stable ground reflectors. Snowcover limits the use to summer time image acquisition.3 Case Histories of Geodynamic ProcessesPlate movementsThe mid-Atlantic plate boundary runs through Iceland, from the tip of the ReykjanesPeninsula in the SW to the Axarfjörður Bay in the NE (Figure 1). Western Iceland is thusmostly a part of the North America Plate and Eastern Iceland is sitting firmly on the EurasiaPlate. The pole of relative rotation between these plates is located in NE-Siberia at 62.4°Nand 135.8°E, and the relative rotation speed is 0.21° per million years according to the Nuvel-1A model of plate motions (DeMets et al. 1994). Holding the North America Plate fixed thisgives a plate velocity vector of 18.2 mm/year in a direction of 105° for Central Iceland,slightly faster and more easterly for South Iceland, slightly slower and more southerly forNorth Iceland. This velocity is valid for the last few millions of years, the time scale of themagnetic and structural data used to constrain the Nuvel-1A model. GPS-data from thecontinuously recording stations (Figure 2) give results that are consistent with the Nuvel-1Avelocity (Geirsson et al., submitted 2005), also preliminary results of measurements of thecountry-wide ISNET network in 1993 and 2004 (Geirsson et al. 2005). This demonstrates thatthe plate movements are consistent on time scales ranging between years and million years.An example of the constant rate of movements is shown in Figure 2, the time series 1999-2005 for the continuous GPS-station at Höfn in SE-Iceland. When the annual cycle ofuncertain origin and the co-seismic effect of the June 2000 earthquakes in South Iceland onthe reference station in Reykjavík have been removed, the graphs shows virtually straightlines. The slopes of the lines give an eastward component of 22.1 mm/year and a soutwardrate of 3.9 mm/year. The vector therefore has a magnitude of 22.4 mm/year and a directon of100° in reasonable agreement with the Nuvel-1A values.Plate boundary deformationBetween the major plates there is a zone of deformation where the crustal movements aredifferent from that of the plates. The width of this deformation zone is somewhat variable. InNorthern Iceland it is about 100 km wide and coincides more or less with the zone ofHolocene volcanism and fissuring. In Southern Iceland the plate boundary has two branchesand a micro-plate can be defined between them, the Hreppar Microplate. The southernboundary of the Hreppar Plate is marked by the South Iceland Seismic Zone where large,strike-slip earthquakes occur.42

Geodynamic Signals Detected by Geodetic Methods in IslandFigure 2: Results of continuous GPS-stations shown as average horizontal displacement vectors with respect to areference station in Reykjavík. Also shown are the time-series for horizontal components of the station at Höfnin SE-Iceland. Note the good fit to the Nuvel-1A model for the stations located on the Eurasia Plate, particularlyRHOF and VMEY.It has been a matter of considerable debate how the plate movements in South Iceland arepartitioned between the two parallel rift zones, the Western Volcanic Zone and the EasternVolcanic Zone. It is generally assumed that the two zones are the expression of a ridge jump,i.e. that the WVZ is a dying rift that is being replaced by the currently much more active EVZ(e.g. Einarsson 1991a). The question is whether the ridge jump occurs by rift propagation, i.e.the EVZ propagating towards the SW while the WVZ recedes, or by activity alternatingbetween the rifts (Sigmundsson et al. 1995) and the whole WVZ gradually becoming lessactive. The lack of evidence for rotated structures within the Hreppar Plate seems to supportthe latter hypothesis. Recent GPS-surveys, however, appear to support rotational movementsof the Hreppar Plate (La Femina et al. 2005), which is in favour of the propagating rifthypothesis. The measurements suggest that near the Hengill triple junction as much as 35% ofthe plate movements is taken up by the WVZ. This proportion dies out towards the NE and is43

Páll Einarsson et al.less than 10% in the Langjökull region. This must indicate a counter-clockwise rotation of theHreppar Microplate, considering the lack of evidence for significant internal deformation ofthat plate. The plate boundary deformation zone accumulates strain during time intervalsbetween significant failure events such as rifting episodes or larger earthquakes. Such gradualaccumulation has been documented for the EVZ by Jónsson et al. (1997), across the SISZ(Sigmundsson et al. 1995, Alex et al. 1999, Perlt and Heinert 2000), and along the ReykjanesPeninsula oblique rift by Sturkell et al. (1994) and Hreinsdóttir et al. (2001).Rifting in the divergent zonesThe Krafla rifting episode provided a dramatic demonstration of crustal deformation along adivergent plate boundary where strain had been accumulating for more than two centuries. Itwas a sequence of magmatic and tectonic events along the plate boundary in N-Iceland,beginning in 1974 and lasting until 1989. It was accompanied by the largest earthquakesequence so far recorded along the divergent plate boundaries of the Atlantic (Einarsson1986). The events took place mainly within the Krafla volcanic system between latitudes of65º34’N and 66º18’N (Figure 1). The volcanic system consists of a central volcano withassociated fissure swarms that extend along the plate boundary perpendicular to the plateseparation vector. During most of the episode, magma apparently ascended from depth andaccumulated in the magma chamber at about 3 km depth beneath the central volcano (e.g.Tryggvason 1980; Ewart et al. 1991). The inflation periods were punctuated by suddendeflation events lasting from several hours to 3 months when the walls of the chamber werebreached and magma was injected laterally into the adjacent fissure swarm wheresubsequently large-scale rifting took place. Rifting, fissuring and graben subsidence tookplace in the fissure swarm but the flanks were uplifted and compressed laterally (e.g.Sigurdsson 1980, Torge and Kanngiesser 1980, Kanngiesser 1983). A total of about 20discrete rifting events were identified, each one affecting only a portion of the fissure system(Björnsson et al. 1979; Tryggvason 1980; Einarsson 1991a,b). Subsidence within the Kraflacaldera was concurrent with rifting and widening of segments of the Krafla fissure swarm(Björnsson et al. 1977, 1979; Tryggvason 1980, 1984, 1994). Early events were primarilyassociated with subsurface movements of magma and little or no lava extrusion. Later in thesequence most of the magma removed from the magma chamber reached the surface infissure eruptions lasting from 5 to 14 days. Maximum cumulative extension of 8-9 m wasmeasured across the fissure swarm slightly north of the Krafla volcano. A large segment ofthe plate boundary was affected by the Krafla events, extending at least 20 km south of Kraflaand 70 km north of the volcano, at least to the rift-transform intersection in Axarfjörður(Figure 1).Co-seismic displacementsIn the transform zones of South and North Iceland the plate boundary is sub-parallel to theplate movement vector. The accumulated strain in these zones is released in large, strike-slipearthquakes, as large as magnitude 7, that take place at intervals of decades to centuries(Einarsson 1991a). The South Iceland Seismic Zone was hit by a series of earthquakes in June2000, two of which caused considerable damage (Stefánsson et al. 2003). The earthquakesfollow a pattern of large historic earthquakes in this zone where sequences of large eventshave occurred at intervals ranging from 45 to 112 years (Einarsson et al. 1981). The sequencebegan on June 17 with a magnitude 6.5 event in the eastern part of the zone. This immediatelytriggered a flurry of activity along at least a 90 km-long stretch of the plate boundary to thewest. This activity included three events with magnitudes larger than 5 on the ReykjanesPeninsula oblique rift (Clifton et al. 2003, Pagli et al. 2003, Árnadóttir et al. 2004). A secondmainshock, also of magnitude 6.5, occurred about 20 km west of the first one on June 21.44

Geodynamic Signals Detected by Geodetic Methods in IslandThe mainshocks of the sequence occurred on N-S striking faults, transverse to the zone itself.The sense of faulting was right-lateral strike-slip conforming to the model of “bookshelffaulting” for the South Iceland Seismic Zone (e.g. Einarsson et al. 1981). According to themodel the left-lateral transform motion across the zone is accomplished by right-lateralmotion along numerous parallel transverse faults and rotation of the blocks between them. Itwas furthermore demonstrated that bookshelf faulting continues to the west, along theReykjanes Peninsula oblique rift (Árnadóttir et al. 2004). One of the events of the sequencehas the characteristics of a “slow earthquake”, i.e. the radiation of seismic waves iscomparatively weak for the amount of faulting observed by InSAR or GPS.The two largest events of the sequence occurred on pre-existing faults and were accompaniedby surface ruptures consisting primarily of en echelon tension gashes and push-up structures(Clifton and Einarsson 2005). The main zones of rupture were about 15 km long, andcoincided with the epicentral distributions of aftershocks. Fault displacements were of theorder of 0.1-1 m at the surface. Faulting along conjugate, left-lateral strike-slip faults alsooccurred, but was less pronounced than that of the main rupture zones.The co-seismic displacement field of the sequence of earthquakes in the SISZ was capturedby InSAR and GPS-measurements (Pedersen et al. 2001, Árnadóttir et al. 2001). Thegeodetic data (Figure 3) were used to invert for the optimal fault geometries and slipdistribution for the two main shocks (Pedersen et al. 2003). According to these modelsfaulting extends from the surface to a depth of 10 km in both events. Maximum displacementsare 2.6 m and 2.9 m, respectively.Post-rifting and post-seismic displacementsThe rifting during the Krafla volcano-tectonic events and the co-seismic displacements duringthe South Iceland earthquakes of June 2000 can be modeled as elastic reaction to the failure ofthe elastic part of the crust under stress that had accumulated in the plate boundary regionduring the previous decades and centuries. The elastic part of the crust appears to be about 10-15 km thick and lies on top of a viscous or visco-elastic material that comprises the lowercrust and the upper mantle. The sudden stress change in the elastic layer leads to increasedstress and viscous reaction in the underlying layer which again induces movements in theelastic surface layer. These movements decay with time in an exponential way. The timeconstantof this decay is directly dependent on the viscosity of the underlying layer. Suchpost-rifting movements have been measured after the Krafla events (Jahn et al. 1994, Foulgeret al. 1992, Heki et al. 1993, Völksen and Seeber 1998, Völksen 2000). Depending on theassumed model parameters such as thickness of the elastic and viscous layers, viscosity valuesin the range 0.3-30 x 10 18 Pa s are obtained.Post-seismic deformation was observed in the SISZ on two spatio-temporal scales followingthe June 2000 earthquake sequence. A rapidly decaying deformation transient, localizedaround the two main shock faults, was captured by several radar interferograms. This signalhas been explained by poro-elastic rebound due to post-earthquake pore-pressure changes(Jónsson et al. 2003). In contrast, the year-scale deformation observed by campaign andcontinuous GPS can be explained by either afterslip at 8-14 km depth or visco-elasticrelaxation of the lower crust and upper mantle in response to the co-seismic stress changes,suggesting viscosities of 0.3-1 x 10 19 Pa s (Árnadóttir et al. 2005).45

Páll Einarsson et al.Figure 3: Horizontal GPS displacements and co-seismic interferograms draped on a radar amplitude image,showing ground deformation associated with the earthquakes in 2000, on June 17 and June 21. Oneinterferometric fringe corresponds to 2.83 cm of range change in the line-of-sight direction. Incoherent areas aremasked. The difference in areal InSAR coverage between Figure A and B is due to utilization of data fromdifferent track frames. Mapped ground ruptures (Clifton and Einarsson 2005) are shown in black, and rivers andseashore are outlined in grey. After Pedersen et al. (2003).46

Geodynamic Signals Detected by Geodetic Methods in IslandVolcano inflation and deflationMany of the volcanoes in Iceland appear to be underlain by shallow-level magma chambers orsemi-permanent bodies of molten material (Sturkell et al. 2006). Activity in the volcanoes isassociated with pressure fluctuations in these bodies. Pressure change in a magma chamberleads to a characteristic displacement field at the surface above and around the chamber,frequently described by the so-called Mogi-model. The model was originally derived for apoint source of pressure or a small sphere of pressure at a specified depth in a half-space.Later it was shown (McTigue 1987) that the model gave a good approximation for the field aslong as (a/d) 5

Páll Einarsson et al.Many opportunities to determine displacement fields around a magma chamber were providedby the Krafla volcano-tectonic episode in 1975-1989, both inflationary and deflationary fields.The magma chamber appeared to be stationary throughout the period. Different inflation anddeflation episodes gave depths in a narrow range around 3 km (e.g. Björnsson et al. 1979,Tryggvason 1980, 1994a, Ewart et al. 1991, Árnadóttir et al. 1998).A large increase in seismic activity near the Hengill triple junction in SW-Iceland in 1994-1998 was associated with uplift of up to 10 centimeters over a wide area as shown by GPSmeasurements,repeated leveling and InSAR. The uplift was interpreted as the result ofmagma injection into the crust at about 7 km depth below the Hrómundartindur volcanicsystem (Sigmundsson et al. 1997, Feigl et al. 2000). The inflation was associated with surfacefracturing (Clifton et al. 2002), damaging earthquakes (about magnitude 5), and increasedthermal activity, but it stopped without an eruption.The Grímsvötn volcano is located near the center of the Iceland hotspot and is almost totallycovered by the Vatnajökull ice cap, which limits the use of geodetic methods to monitor itsactivity. Repeated GPS-measurements on the only useable nunatak, Grímsfjall on the calderarim, have given very characteristic time changes that can be interpreted with the help of theMogi-model (Sturkell et al. 2003a, 2005a). The volcano deflated during the eruptions of 1998and 2004, and inflated in the time period between the eruptions. The inflation rate was used togive a long-term forecast for the 2004 eruption (Sturkell et al. 2003a, Vogfjörð et al. 2005).Similarly, the present reinflation rate of the volcano indicates that an eruption is to beexpected within a few years.The Hekla volcano has erupted a few times within the time period of precise geodesy, i.e.1970, 1980-81, 1991 and 2000. Several attempts have been made to determine a depth to amagma chamber feeding the eruptions but the results are vague and partly contradictory. Tiltmeasurements have indicated an inflation center west of the volcano (Tryggvason 1994b),EDM-trilateration in association with the 1980-81 activity gave a chamber at 8 km depth(Kjartansson and Grönvold 1983), and GPS-measurements during the 1991 eruption gave arather uncertain depth between 2 and 11 km (Sigmundsson et al. 1992). A joint interpretationof GPS-data, InSAR, tilt, and volumetric strain changes associated with the 2000 eruptionindicates a broad deformation field from a rather deep-seated magma chamber at 11 km depth,disturbed by irregular, local sources, most likely due to loading of the surface by lava flows(Sturkell et al. 2005). Current tilt changes indicate that the volcano is re-inflating and that itmay have reached the pre-eruption inflation level already.The Askja volcano in the Northern Volcanic Zone has been the target of many geodeticstudies since it latest eruption in 1961. Eysteinn Tryggvason, the pioneer of volcano geodesyin Iceland, installed a levelling profile in 1966-68 which has been remeasured and extendedmany times since then. During the first years the volcano was inflating, but about 1973 thetrend was reversed to deflation. The deflation was rapid in the beginning but the rate has beendecreasing in an exponential manner with a decay constant of 39 years (Sturkell et al.,submitted 2004). The geodetic data have been modelled and interpreted in terms of a singleMogi-type pressure source located close to the centre of the Askja main caldera (Tryggvason1989; Rymer and Tryggvason 1993; Sturkell and Sigmundsson 2000). All these authorsplaced the point source at 1.5 to 3.5 km depth. This model accounts for most of the observeddisplacements in the main caldera and its immediate vicinity. At a greater distance, however,displacements observed with GPS do not show the same good fit. A more elaborate model ispresented by Sturkell et al. (submitted 2004) who invoke two Mogi sources to account for thefar field displacements. This model was also applied to the results of repeated micro-gravity48

Geodynamic Signals Detected by Geodetic Methods in Islandstudies for the period 1988-2003. A sub-surface mass decrease of 1.6 x 10 11 kg is derived (deZeeuw et al., submitted 2004) indicating that magma drainage is an important contributor tothe sub-surface mass decrease. The geometry of the deeper pressure source is not wellresolved but it is suggested that it is at 16 km depth and is elongated along the axis of thefissure swarm. This is confirmed in a general way by a study by Pagli et al. (2005), who addthe constraints of InSAR data and use ellipsoidal models in an inversion for depth andchamber size (Figure 5). They explain most of the displacement field by an ellipsoidalpressure source at about 3 km depth, but an elongated zone of subsidence along the associatedfissure swarm is ascribed to a deeper source of decreasing pressure.Figure 5: Displacements in the surrounding of the Askja volcano measured by InSAR (left panel) and GPS (bothpanels). Nested calderas of Askja are shown with thin lines. InSAR fringes in the caldera region show deflationfrom a shallow source beneath the caldera. Elongated fringes outside the caldera are extended along the fissureswarm of the volcano and show subsidence, possibly due to extension across the plate boundary. GPS-vectorsshow horizontal contraction towards the volcano. After Pagli et al. (2005).IntrusionsTwo separate intrusion events occurred beneath the Eyjafjalljökull volcano in South Iceland,the first one in 1994 and the second one in 1999. Both events began with increased earthquakeactivity beneath the NE flank of the volcano followed by inflation centered on the S flank.Ground deformation indicating inflation in 1994 was initially discovered by tilt and GPSgeodesyand appears to have occurred mostly during the peak of the seismic activity (Sturkellet al. 2003b). Inflation was further confirmed by InSAR, with a center beneath the S-flank at4.5 km depth (Pedersen and Sigmundsson 2004). In 1999 the InSAR data showed that theinflation bulge was displaced slightly S with respect to the 1994 inflation bulge (Pedersen andSigmundsson 2005), and a deeper source is required to fit the data (Figure 6). The InSAR datafavour a sill-shaped intrusive body for both events. Estimated volume of intruded magma in1994 is about 17 x 10 6 m 3 and in 1999 it is 30 x 10 6 m 3 .49

Páll Einarsson et al.Figure 6: Ground displacements recorded by InSAR in the vicinity of the Eyjafjallajökull volcano betweenAugust 01, 1997 and Sept. 29, 2000 including the intrusion event of 1999. One interferometric fringecorresponds to 2.83 cm of range change in the line-of-sight direction. Incoherent areas are masked. The outlineof the Eyjafjallajökull glacier covering the summit region of the volcano is shown with a black line, andlocations of GPS stations in the area are shown with stars. After Pedersen and Sigmundsson (2005).Glacial isostasy and uplift around VatnajökullThe glaciers in Iceland provide variable loads on the crust that result in isostatic adjustments.By measuring the crustal response to the load changes one can estimate the thickness of theelastic upper crust and the viscosity in the underlying visco-elastic layer. Deglaciation inIceland at the end of the Weichselian glaciation about 10000 years BP, was associated withrapid glacial rebound, being completed in only about 1000 years in coastal areas. Thisexceptionally fast postglacial rebound has been modeled to argue for viscosity under Icelandof the order of 10 19 Pa s or less (Sigmundsson 1991). The low viscosity results in a rapidresponse of the Earth to contemporary changes in ice volume. In the last century, the icevolume of the Vatnajökull glacier has significantly decreased. Ongoing uplift aroundVatnajökull is reported by several geodetic studies. Lake leveling measurements at LakeLangisjór at Western edge of Vatnajökull were performed in 1959-1991 (Sigmundsson andEinarsson 1992). Measurements show uplift rate of about 4 mm/yr between benchmarksspaced 15 km perpendicular to the ice edge. In 1991 a GPS network of ten points was firstmeasured around the Southeastern edge of Vatnajökull (Einarsson et al. 1996). Gravitymeasurements were also conducted at all GPS stations and annually repeated until 2000,except in 1994 (Jacoby et al. 2001). Gravity changes between 1991 and 2000 for points close50

Geodynamic Signals Detected by Geodetic Methods in Islandto the ice cap are consistent with uplift rates up to 20 mm/year. In 1992, 1996 and in 1999 theoriginal 1991 GPS network was remeasured and eleven additional points were included in thenetwork in 1992 (Sjöberg et al. 2000 and 2004). The uplift rate 1992-1999 was estimated tobe about 5-19 mm/yr, decaying radially from the center of the ice cap (Sjöberg et al 2004).Thoma and Wolf (2001) used the lake level measurements of Lake Langisjór from 1959 to1991 and GPS measurements from 1992 to 1996 (Sjöberg et al. 2000) to constrain therheology in Iceland. The authors use a compressible, self-gravitating, spherical Earth modelwith Maxwell viscoelasticity and an elliptic ice load. They consider two different ice thinningmodels. Modeling results suggest a lower crust/upper mantle viscosity between 7×10 16 -3×10 18Pa s and a thickness of the elastic crust between 10-20 km. Similar results were reported bySjöberg et al. (2004) who processed GPS measurements in 1992, 1996 and 1999. Theyconclude that the vertical GPS velocities can be fit by assuming an elastic thickness of thecrust on the order of 10-20 km and a viscosity perhaps as low as 1×10 17 Pa s.Jacoby et al. (2001) measured gravity changes between 1991 and 2000 and compared them tothe model by Sigmundsson and Einarsson (1992). Results suggest a lower crust/upper mantleviscosity on the order of 10 18 Pa s and a thickness of the elastic crust of about 10 km.In 1996 and in 2003 a GPS network of 15 points was measured around the south edge ofVatnajökull. GPS vertical velocities around the ice cap vary between 7-25 mm/yr (Figure 7).These GPS and lake leveling measurements have been modeled using the Finite ElementMethod (FEM). Results indicate a thickness of the elastic crust of 10-20 km and a viscosity ofthe lower crust/upper mantle of 3-8 x10 18 Pa s (Pagli et al., in preparation). Knowledge of theEarth structure allows us to predict uplift around Vatnajökull in the next decades.Figure 7: Vertical displacements of GPS-points near the edge of Vatnajökull glacier during 1996-2003 (inITRF00 reference frame). The uplift rate is highest near the glacier and decreases with increasing distance fromits edge. From Pagli et al. (in prep. 2005).51

Páll Einarsson et al.Mass movements suggested in the roots of the Iceland hotspotPreliminary results of the 2004 re-measurements of the ISNET geodetic net of Iceland suggestthat most points in the interior of Iceland are being uplifted at a rate of 10-20 mm/year(Geirsson et al. 2005). Part of this can be explained by isostatic uplift following the recentreduction in ice load of the present glaciers. The uplift appears to be too widespread, however,for all of it to be modeled by this process. The residual signal may have to be explained byadditional processes taking place in the lower crust or the upper mantle. In fact, massmovements and pressure connection at these levels have been suggested to explain apparentcorrelation in time between activity at different volcanoes (e.g. Einarsson 1991, Tryggvason1989, Sturkell et al. 2006). For further modeling of deep mass movements the importance ofprecise gravity measurements becomes apparent. Absolute gravity measurements wereinitiated in Iceland in 1987-1988 (Torge et al. 1992) at 5 points, one in Reykjavík and 4 inNorth Iceland. Additional measurements were done in 1997 by IfAG (now Bundesamt fürKartographie und Geodäsie) at 7 points, including Reykjavík and Höfn, and 2 points in theinterior of Iceland. Further measurements were done at Reykjavík and Höfn by the FinnishGeodetic Institute. Repetition of these measurements and future expansion of this absolutegravimetry network is likely to contribute to better understanding of the dynamics of theIceland hotspot.4 ConclusionsIceland has been, and still is, an important test ground for geodynamic models, both on aglobal and local scale. New geodetic techniques have been used and tested in Iceland for thispurpose for decades.The basic assumption of the plate tectonics theory has been verified, i.e. that the surface of theEarth is divided into plates with insignificant internal deformation separated by plateboundary deformation zones. It has been shown that the global model of plate movementsapplies well in Iceland. Measured plate movements of the last few years conform with theglobally determined plate movements of the last few million years. Iceland sits on two of themajor plates, the Eurasia and North America Plates. A small plate fragment, the HrepparMicroplate, is defined between the two volcanic rift zones in South Iceland. Furthermicroplates or crustal blocks within the Tjörnes Fracture Zone remain to be defined.Crustal dynamics models with an elastic plate on top of a viscous layer or half-space havebeen successfully used to explain crustal movements measured in conjunction with riftingevents in the divergent plate boundary zones and earthquakes in the transform zones, both theimmediate elastic response and the post-event movements. The thickness of the elastic plate isin the range 10-15 km and the viscosity in the underlying layer is in the range 0.3-30 x 10 18 Pas. Similar models have been used in interpreting glacio-isostatic movements observed aroundthe Vatnajökull ice cap in response to reduced ice load in the last century. It is important tomaintain some of the established time series of movements in order to refine these models,e.g. by introducing further layering.Crustal deformation measurements provide some of the most useful data for the monitoring ofactive volcanoes. Pressure changes in the magma systems of volcanoes are reflected bycrustal movements that can be measured. Mogi-type models, i.e. pressure point sources in anelastic half-space, have been very successfully applied to deformation fields measured aroundactive volcanoes. These models are relatively insensitive to shape and size of the magmachambers where the pressure changes occur. They do, however, reveal whether the magmapressure is increasing or decreasing, and give an indication of the depth to the magma52

Geodynamic Signals Detected by Geodetic Methods in Islandchamber. At the present time the magma pressure is increasing beneath the three most activeIcelandic volcanoes, Grímsvötn, Hekla and Katla. Pressure is decreasing in the shallow-levelmagma chambers of Askja and Krafla.Geodetic data are becoming available that have better continuity in space (InSAR) and in time(CGPS). These data demand refined models for their interpretation. Fault models with nonuniformslip distributions and magma chambers with ellipsoidal shapes are applied withconsiderable success in inversion algorithms.A multitude of geodynamic problems are still unsolved and exciting geodetic experiments canbe carried out in Iceland. These can benefit from ever more sophisticated geodeticmeasurements, with improved spatial and temporal resolution. Further densification of thenetwork of continuous GPS stations may be particularly rewarding, augmented by broad-bandseismic stations and absolute gravimetry. Multi-national cooperative geodetic projects haveturned out to be very fruitful, such as collaborative work with German geodesists. Continuedand improved German-Iceland collaboration in geophysical geodesy may therefore be afruitful way to improve the understanding of Iceland geodynamics.ReferencesAlex, N., P. Einarsson, M. Heinert, W. Niemeyer, B. Ritter, F. Sigmundson, St. Willgalis (1999).GPS-Messkampagne 1995 zur Bestimmung von Deformationen der Erdkruste in Sudwestisland.Zeitschrift fur Vermessungswesen, 124, 347-361.Árnadóttir, Th., Sigmundsson, F. and Delaney, P. T. (1998). Sources of crustal deformation associatedwith the Krafla, Iceland, eruption of September 1984, Geophys. Res. Lett., 25, 1043-1046.Árnadóttir, Th., S. Hreinsdóttir, G. Guðmundsson, P. Einarsson, M. Heinert, C. Völksen (2001).Crustal deformation measured by GPS in the South Iceland Seismic Zone due to two largeearthquakes in June 2000. Geophys. Res. Lett., 28, 4031-4033.Árnadóttir, Th., H. Geirsson, P. Einarsson (2004). Coseismic stress changes and crustal deformationon the Reykjanes Peninsula due to triggered earthquakes on June 17, 2000. J. Geophys. Res.,109, B09307, doi:10.1029/2004JB003130.Árnadóttir, Th., S. Jónsson, F.F. Pollitz, W. Jiang and K.L. Feigl (2005). Postseismic deformationfollowing the June 2000 earthquake sequence in the south Iceland seismic zone, J. Geophys.Res., 110, B12308, doi:10.1029/2005JB003701.Björnsson, A., Saemundsson, K., Einarsson, P., Tryggvason, E. and Grönvold, K. (1977). Currentrifting episode in north Iceland. Nature, 266, 318-323.Björnsson, A., Johnsen, G., Sigurðsson, S., Thorbergsson, G. and Tryggvason, E. (1979). Rifting ofthe plate boundary in North Iceland 1975-1978. J. Geophys. Res., 84: 3029-3038, Brander, J. L.,R. G. Mason, R. W. Calvert. Precise distance measurements in Iceland. Tectonophysics, 31,193-206.Clifton, A., F. Sigmundsson, K. Feigl, G. Gunnarsson, Th. Árnadóttir (2002). Surface effects offaulting and deformation resulting from magma accumulation at the Hengill triple junction, SWIceland, 1994-1998. J. Volcanol., Geothermal Res., 115, 233-255.Clifton, A.E., C. Pagli, J. F. Jónsdóttir, K. Eythórsdóttir and K. Vogfjörd (2003). Surface effects oftriggered fault slip on Reykjanes Peninsula, SW Iceland, Tectonophysics, 369, 145-154.Clifton, A., P. Einarsson (2005). Styles of surface rupture accompanying the June 17 and 21, 2000earthquakes in the South Iceland Seismic Zone. Tectonophysics, 396, 141-159.Decker, R. W., P. Einarsson and P. A. Mohr (1971). Rifting in Iceland: New geodetic data, Science173, 530-533.DeMets, R. G. Gordon, D. F. Argus, S. Stein (1994). Effect of recent revisions to the geomagneticreversal time scale on estimates of current plate motions. Geophys. Res. Lett., 21, 2191-2194.53

Páll Einarsson et al.Einarsson, P. (1986). Seismicity along the eastern margin of the North American Plate. in Vogt, P.R.,and Tucholke, B.E., eds., The Geology of North America, Vol. M, The Western North AtlanticRegion: Geological Society of America, p. 99-116.Einarsson, P. (1991a). Earthquakes and present-day tectonism in Iceland. Tectonophysics, 189, 261-279.Einarsson, P. (1991b). The Krafla rifting episode 1975-1989, in Náttúra Mývatns, (The Nature of lakeMývatn), edited by A. Gardarsson, and Á. Einarsson, pp. 97-139, Icelandic Nature ScienceSociety, Reykjavíc.Einarsson, P. (2001). Structure and evolution of the Iceland hotspot. Deutsche GeophysikalischeGesellschaft, Mitteilungen, 1/2001, 11-14.Einarsson, P., S. Björnsson, G. Foulger, R. Stefánsson and Þ. Skaftadóttir (1981). Seismicity pattern inthe South Iceland seismic zone, in: Earthquake Prediction - An International Review (eds. D.Simpson and P. Richards). American Geophys. Union, Maurice Ewing Series 4, 141-151.Einarsson, P., and K. Sæmundsson (1987). Earthquake epicenters 1982-1985 and volcanic systems inIceland: A map, in Í hlutarins eðli, Festschrift for Þorbjörn Sigurgeirsson,, edited by Sigfússon,Menningarsjóður, Reykjavík.Einarsson, P., F. Sigmundsson, M. Hofton, G. Foulger, W. Jacoby (1996). An experiment in glacioisostacynear Vatnajökull, Iceland, 1991. Jökull, 44, 29-39.Ewart, A., B. Voight, A. Björnsson (1991). Elastic deformation models of Krafla Volcano, Iceland, forthe decade 1975 trough 1985. Bull. Volcanol., 53, 436-459.Feigl, K., J. Gasperi, F. Sigmundsson, A. Rigo (2000). Crustal deformation near Hengill volcano,Iceland 1993-1998: Coupling between volcanism and faulting inferred from elastic modelling ofsatellite radar interferograms. J. Geophys. Res., 105, 25655-25670.Foulger, G., R. Bilham, W. J. Morgan, and P. Einarsson (1987). The Iceland GPS geodetic fieldcampaign 1986. EOS, 68, 1809.Foulger, G. R., C.-H. Jahn, G. Seeber, P. Einarsson, B. R. Julian, K. Heki (1992). Post-rifting stressrelaxation at the divergent plate boundary in Northeast Iceland, Nature, 358, 488-490.Foulger, G., G. Beutler, R. Bilham, P. Einarsson, S. Fankhauser, W. Gurtner, U. Hugentobler, W. J.Morgan, M. Rothacher, G. Thorbergsson, and U. Wild (1993). The Iceland 1986 GPS geodeticsurvey: Tectonic goals and data processing results, Bull. Geodetica, 67, 148-172.Geirsson, H., Th. Árnadóttir, C. Völksen, W. Jiang, E. Sturkell, T. Villemin, P. Einarsson, F.Sigmundsson, R. Stefánsson (2005). Current plate movements across the Mid-Atlantic Ridgedetermined from 5 years of continuous GPS measurements in Iceland. J. Geophys. Res.,submitted 2005.Geirsson, H., Th. Árnadóttir,E. Sturkell, W. Jiang, M. Rennen, C. Völksen, C. Pagli, T. Sigurdsson, T.Theodorsson, J. Erlingsson, P. Einarsson, F. Sigmundsson (2005). Crustal deformation inIceland derived from the nation-wide 1993 and 2004 ISNET campaigns, Eos Trans. AGU,86(52), Fall Meet. Suppl., Abstract G21B-1275.Gerke, K. (1974). Crustal movements in the Myvatn- and the Thingvallavatn-Area, both horizontaland vertical. In: L. Kristjánsson (ed.), Geodynamics of Iceland and the North Atlantic Area. p.263-275, D. Reidel Publ. Comp., Dordrecht, Holland.Heki, K., G. R. Foulger, B. R. Julian, C.-H. Jahn (1993). Plate kinematics near divergent boundaries:Geophysical implications of post-tectonic crustal deformation in NE-Iceland detected using theGlobal Positioning System. J. Geophys. Res., 98, 14279-14297.Hreinsdóttir, S., P. Einarsson, F. Sigmundsson (2001). Crustal deformation at the oblique spreadingReykjanes Peninsula, SW Iceland: GPS measurements from 1993 to 1998. J. Geophys. Res.,106, 13,803-13,816.Jacoby, W. R., Bürger, S., Smilde, P. and Wallner, H. (2001). Temporal gravity variations observed inSE Iceland, InterRidge News, 10(1), 52-55.Jahn, C.-H. (1992). Untersuchungen über den Einsatz des Global Positioning System (GPS) zumNachweis rezenter Krustendeformationen im Spaltengebiet Nordost-Islands. WissenschaftlicheArbeiten der Fachrichtung Vermessungswesen der Universität Hannover, Nr. 182, Hannover.Jahn, C.-H., G. Seeber, G.R. Foulger and P. Einarsson (1994). GPS epoch measurements spanningthe mid-Atlantic plate boundary in northern Iceland 1987-1990. In: Gravimetry and spacetechniques applied to geodynamics and ocean dynamics. Geophys. Monogr. Ser., vol. 82 (ed. B.E. Schutz et al.), AGU, Washington, p.109-123.54

Geodynamic Signals Detected by Geodetic Methods in IslandJohnsen, G. V., A. Björnsson, and S. Sigurðsson (1980). Gravity and elevation changes caused bymagma movements beneath the Krafla caldera, Northern Iceland. J. Geophys., 47, 132-140.Jónsson, S., P. Einarsson, F. Sigmundsson (1997). Extension across a divergent plate boundary, theEastern Volcanic Rift Zone, south Iceland, 1967-1994, observed with GPS and electronicdistance measurements. J. Geophys. Res., 102, 11,913-11,930.Jónsson, S., P. Segall, R. Pedersen, and G. Björnsson (2003). Post-earthquake ground movementscorrelated to pore-pressure transients, Nature, 424, 179-183.Kanngiesser, E. (1983). Vertical component of ground deformation in north Iceland. AnnalesGeophysicae, 1, 321-328.Kjartansson E, K. Grönvold (1983). Location of a magma reservoir beneath Hekla volcano, Iceland.Nature, 301, 139-141.La Femina, P.C., T.H. Dixon, R. Malservisi, Th. Árnadóttir, E. Sturkell, F. Sigmundsson, and P.Einarsson (2005). Geodetic GPS measurements in south Iceland: Strain accumulation andpartitioning in a propagating ridge system, J. Geophys. Res., 110, B11405,doi:10.1029/2005JB003675.Massonnet, D. and Sigmundsson, F. (2000). Remote sensing of volcano deformation by radarinterferometry from various satellites, In: Remote sensing of volcanoes, Am. Geophys. UnionGeophysical Monograph, 116, 207-221.McTigue, D. F. (1987). Elastic stress and deformation near a finite spherical magma body: Resolutionof the point source paradox. J. Geophys. Res., 92, 12,931-12,940.Niemczyk, O. (1943). Spalten auf Island. Verlag Konrad Wittwer, Stuttgart, 180 pp.Pagli, C., R. Pedersen, F. Sigmundsson, K. Feigl (2003). Triggered fault slip on June 17, 2000 on theReykjanes Peninsula, SW Iceland, captured by radar interferometry. Geophys. Res. Lett., 30,doi:10.1029/2002GL015310.Pagli, C. F. Sigmundsson, T. Árnadóttir, P. Einarsson and E. Sturkell (2005). Deflation of the Askjavolcanic system: constraints on the deformation source from combined inversion of satelliteradar interferograms and GPS measurements. J. Volc. Geothermal Res., doi:10.1016/j.jvolgeores.2005.09.014. Online 2005.Pagli, C., F. Sigmundsson, H. Geirsson, E. Sturkell, B. Lund, P. Einarsson, T. Árnadóttir (2005).Current and future glacio-isostatic deformation around Vatnajökull ice cap, Iceland. In prep.2005.Pedersen, R., F. Sigmundsson, K. Feigl, Th. Árnadóttir (2001). Coseismic interferograms of twoMs=6.6 in the South Iceland Seismic Zone, June 2000. Geophys. Res. Lett., 28, 3341-3344.Pedersen, R., S. Jónsson, Th. Árnadóttir, F. Sigmundsson, K. Feigl (2003). Fault slip distribution oftwo June 2000 Mw 6.4 earthquakes in South Iceland estimated from joint inversion of InSARand GPS measurements. Earth Planet. Sci. Lett., 213, 487-502.Pedersen, R. and F. Sigmundsson (2004). InSAR based sill model links spatially offset areas ofdeformation and seismicity for the 1994 unrest episode at Eyjafjallajökull volcano, Iceland.Geophys. Res. Lett., 31, L14610, doi:10.1029/2004GL020368.Pedersen, R. and F. Sigmundsson (2005). Temporal development of the 1999 intrusive episode in theEyjafjallajökull volcano, Iceland, derived from InSAR images. Bull. Volc., doi: 10.1007/s00445-005-0020-y.Perlt, J. and Heinert, M. (2000). Kinematic model of the South Iceland tectonic system, Geophys. J.Int., 142, 1-9.Rymer, H. and Tryggvason, E. (1993). Gravity and elevation changes at Askja, Iceland. Bull. Volc,55(5): 362-371.Seeber, G. (1989). GPS-Messungen in der jungvulkanischen Zone Nordost-Islands. Zeitschrift fürVermessungswesen 114, 39-43.Sigmundsson, F. (1991). Post-glacial rebound and asthenosphere viscosity in Iceland. Geophys. Res.Letters, 18, 1131-1134.Sigmundsson, F. (2006). Iceland Geodynamics, Crustal Deformation and Divergent Plate Tectonics.Springer – Praxis Publishing Ltd, Chichester, UK, 209 pp.Sigmundsson, F., P. Einarsson, R. Bilham (1992). Magma chamber deflation recorded by the GlobalPositioning System: The Hekla 1991 eruption, Geophys. Res. Lett., 19, 1483-1486.Sigmundsson, F., P. Einarsson (1992). Glacio-isostatic crustal movements caused by historical volumechange of the Vatnajökull ice cap, Iceland, Geophys. Res. Lett., 19, 2123-2126.55

Páll Einarsson et al.Sigmundsson, F., P. Einarsson, R. Bilham, E. Sturkell (1995). Rift-transform kinematics in southIceland: Deformation from Global Positioning System measurements, 1986 to 1992. J.Geophys. Res., 100, 6235-6248.Sigmundsson, F., H. Vadon, D. Massonet (1997). Readjustment of the Krafla spreading segment tocrustal rifting measured by Satellite Radar Interferometry. Geophys. Res. Lett., 24, 1843-1846.Sigmundsson, F., P. Einarsson, S. Rögnvaldsson, G. R. Foulger, K. M. Hodgkinson, G. Thorbergsson(1997). The 1994-1995 seismicity and deformation at the Hengill triple junction, Iceland:Triggering of earthquakes by minor magma injection in a zone of horizontal shear stress. J.Geophys. Res., 102, 15,151-15,161.Sigurdsson, O. (1980). Surface deformation of the Krafla fissure swarm in two rifting events. J.Geophys., 47, 154-1159.Sjöberg, L. E., M. Pan, E. Asenjo, S. Erlingsson (2000). Glacial rebound near Vatnajökull, Iceland,studied by GPS campaigns in 1992 and 1996. J. Geodynamics, 29, 63-70.Sjöberg, L., M. Pan, S. Erlingsson, E. Asenjo, K. Árnason (2004). Land uplift near Vatnajökull,Iceland, as observed by GPS in 1992, 1996, and 1999. Geophys. J. Int., 159, 943-948.Stefánsson, R., G. Gudmundsson, and P. Halldórsson (2003). The South Iceland earthquakes 2000 -a challenge for earthquake prediction research. Report VI-R03017, 21 pp, IcelandicMeteorological Office, Reykjavík, Iceland.Sturkell, E., F. Sigmundsson, P. Einarsson, R. Bilham (1994). Strain accumulation 1986-1992 acrossthe Reykjanes Peninsula plate boundary, Iceland, determined from GPS measurements.Geophys. Res. Lett., 21, 125-128.Sturkell, E., Sigmundsson, F. (2000). Continuous deflation of the Askja caldera Iceland, during the1983-1998 non-eruptive period, J. Geophys. Res., 105, 25671-25684.Sturkell, E., P. Einarsson, F. Sigmundsson, S. Hreinsdóttir, H. Geirsson (2003a). Deformation ofGrímsvötn volcano, Iceland: 1998 eruption and subsequent inflation. Geophys. Res. Lett., 30,10.1029/2002GL016460.Sturkell, E., F. Sigmundsson, and P. Einarsson (2003b). Recent unrest of the Eyjafjallajökull and Katlavolcanoes, Iceland. J. Geophys. Res. 108, No. B8, 2369, 10.1029/2001JB00091.Sturkell, E., Sigmundsson, F., Slunga, R. (2004). 1983-2003 decaying rate of deflation at Askjacaldera: Pressure decrease in an extensive magma plumbing system at a spreading plateboundary. Bull. Volc., submitted 2004.Sturkell, E., K. Ágústsson, A. Linde, S. Sacks, P. Einarsson, F. Sigmundsson, H. Geirsson, R.Pedersen and P. La Femina (2005). Geodetic constraints on the magma chamber of the Heklavolcano, Iceland. Eos. Trans. AGU, 86 (48), Fall Meeting Suppl., Abstract V21D-0636.Sturkell, E., P. Einarsson, F. Sigmundsson, H. Geirsson, R. Pedersen, E. Van Dalfsen, A. Linde, S.Sacks, R. Stefánsson (2006). Volcano geodesy and magma dynamics in Iceland. J. Volc.Geothermal Res., 150, 14-34.Thoma, M., and D. Wolf. (2001). Inverting land uplift near Vatnajökull, Iceland, in terms oflithosphere thickness and viscosity stratification, in: M. G. Sideris (ed.), Gravity, Geoid andGeodynamics 2000 (pp. 97-102), Springer Verlag, Berlin.Torge, W. (1989) Schweremessungen in Nordostisland 1938-1987. Zeitschrift für Vermessungswesen114, 44-55.Torge, W., H. Drewes (1977). Gravity variations with time in Northern Iceland 1965-1975. J.Geophys., 43, 771-790.Torge, W., E. Kanngiesser (1980). Gravity and height variations during the present rifting episode inNorthern Iceland. J. Geophys., 47, 125-131.Torge, W., T. Grote, R. H. Röder, M. Schnüll, H.-G. Wenzel (1992). Introduction of AbsoluteGravimetric Techniques into a High-Precision Gravity and Vertical Control System in NorthernIceland. Deutsche Geodätische Kommission B 297, 52 pp.Tryggvason, E. (1980). Subsidence events in the Krafla area, North Iceland, 1975-1979. J. Geophys.,47, 141-153.Tryggvason, E. (1984). Widening of the Krafla fissure swarm during the 1975-1981 volcano-tectonicepisode. Bull. Volcanol., 47, 47-69.Tryggvason, E. (1989). Ground deformation in Askja, Iceland: Its source and possible relation to flowof the mantle plume. J. Volcanol. Geotherm. Res. 39, 61–71.56

Geodynamic Signals Detected by Geodetic Methods in IslandTryggvason, E. (1994a). Surface deformation at the Krafla volcano, North Iceland, 1982-1992, Bull.Volcanol., 56, 98-107.Tryggvason, E. (1994b). Observed ground deformation at Hekla, Iceland prior to and during theeruptions of 1970, 1980-1981 and 1991. J. Volcanol. Geotherm. Res., 61, 281-291.Vadon, H., and F. Sigmundsson (1997). Crustal deformation from 1992 to 1995 at the Mid-AtlanticRidge, Southwest Iceland, mapped by satellite radar interferometry. Science 275, 193-197.Vogfjörð, K., S. Jakobsdóttir, G. B. Guðmundsson, M. J. Roberts, K. Ágústsson, T. Arason, H.Geirsson, S. Karlsdóttir, S. Hjaltadóttir, U. Ólafsdóttir, B. Thorbjarnardóttir, T. Skaftadóttir, E.Sturkell, E. B. Jónasdóttir, G. Hafsteinsson, H. Sveinbjörnsson, R. Stefánsson, T. V. Jónsson(2005). Forecasting and monitoring a subglacial eruption in Iceland. EOS, Transactions,American Geophys. Union, 86, 245.Völksen, C. (2000). Die Nutzung von GPS für die Deformationsanalyse in regionalen Netzen amBeispiel Islands. Wissenschaftlichen Arbeiten der Fachrichtung Vermessungswesen derUniversität Hannover, Nr. 237, Hannover.Völksen, C. and Seeber, G. (1998). Nachweis von rezenten Krustendeformationen in Nordisland mitGPS, Zeitschrift für Vermessungsvesen, 123, 68-75.de Zeeuw-van Dalfsen, E., Rymer, H., Sigmundsson, F., Sturkell, E. (2004). Net gravity decrease atAskja volcano, Iceland: Constraints on processes responsible for continuous caldera deflation,1988-2003. J. Volcanology and Geothermal Research, submitted 2004.57